A conventional solid-state imaging device converts light, which is detected by a pixel unit including a photodiode, into charges, and outputs the charges by transferring the converted charges by a charge transferring unit configured by a Charge Coupled Device (CCD) or a CMOS device. In the solid-state imaging device, a plurality of charge accumulating units extracting charges generated in the pixel unit, accumulating the extracted charges, and transferring the accumulated charges to the charge transferring unit, are formed between the pixel unit and the charge transferring unit.
The respective charge accumulating units are manufactured by forming an N-type impurity layer constituting a photodiode on a surface of a P-type semiconductor substrate such that the N-type impurity layer protrudes from the pixel unit, and then, forming a gate electrode such that the gate electrode overlaps a portion of a section in which the N-type impurity layer protrudes.
In the charge accumulating units manufactured in this manner, a residual image characteristic is dependent on a potential of the pixel unit, and is also dependent on a potential (charge reservoir) formed by the N-type impurity layer overlapping the gate electrode. Different residual image characteristics are required in different products.
However, since it is difficult to visually determine a position of the impurity layer, it is difficult to form the gate electrode based on the position of the impurity layer. Therefore, the position of the formed gate electrode varies. As a result, an area of a section in which the N-type impurity layer and the gate electrode overlap each other varies greatly in each charge accumulating unit, and a depth of the charge reservoir also varies greatly. Hence, a residual image characteristic in each charge accumulating unit varies, and a residual image characteristic in each product also varies.